🏆 For Judges / 3 demo personas
🧑‍🤝‍🧑 Personas · Schema _j 1.2

Three lives. One standard.

Each persona's IPS profile below is a real _j 1.2 JSON payload, identical in shape to what the JemmaPass app reads from a QR scan. The profiles are clinically realistic and used verbatim in the demo video.

🐺
Kurodo
Kurodo
The Belgian pilgrim · 黒人遍路
🇧🇪 Belgian national 🚶‍♂️ Henro pilgrim 🗣️ FR · EN 🩹 Multiple allergies

Background. 58 years old, retired civil servant from Brussels, walking the Shikoku 88-temple pilgrimage for spiritual reasons. He speaks French natively, conversational English, and no Japanese. He carries an EpiPen because of his severe penicillin allergy, with a documented history of anaphylaxis.

Why he matters in the demo. Kurodo is the patient who triggers the killer moment. His IPS profile, embedded in a 500-byte QR worn around his neck, contains a SNOMED-coded penicillin allergy that JemmaPass walks the ATC hierarchy on to flag Augmentin as contraindicated — preventing a fatal anaphylactic reaction during the earthquake aftermath.

📄 Kurodo's _j 1.2 profile

Raw: ~1100 bytes Wire: ~550 bytes Compression: deflate-raw
{
  "_j": "1.2",
  "p": {
    "gn": "Kurodo",
    "fn": "Henro",
    "gs": "M",
    "bd": "1968-04-12",
    "nat": "BE",
    "bt": "A+",
    "ct": ["+32-470-XX-XX-XX"]
  },
  "al": [
    {
      "c": "91936005",
      "s": "H",
      "st": "A",
      "d": "Allergy to penicillin",
      "m": "Anaphylactic shock 2019-03",
      "d_display": "ペニシリンアレルギー · Allergie à la pénicilline"
    },
    {
      "c": "232347008",
      "s": "H",
      "st": "A",
      "d": "Allergy to fish",
      "m": "Urticaria + tongue swelling",
      "d_display": "魚アレルギー · Allergie au poisson"
    },
    {
      "c": "419263009",
      "s": "L",
      "st": "A",
      "d": "Allergy to tree pollen",
      "d_display": "花粉症 · Pollinose"
    }
  ],
  "md": [
    {
      "c": "C09AA02",
      "t": "Enalapril 10mg",
      "r": "O",
      "v": "1",
      "u": "tab",
      "rs": "Daily",
      "rc": "I10",
      "d_display": "Hypertension treatment · 高血圧症治療"
    },
    {
      "c": "B01AC06",
      "t": "Aspirin 100mg",
      "r": "O",
      "v": "1",
      "u": "tab",
      "rs": "Daily",
      "rc": "I25",
      "d_display": "Cardioprotection · 心保護"
    }
  ],
  "cn": [
    {
      "c": "I10",
      "d": "Essential hypertension",
      "st": "A",
      "d_display": "高血圧症 · Hypertension"
    }
  ]
}
Why this matters: Notice the SNOMED CT code 91936005 for penicillin allergy. When Kamekichi scans Augmentin (ATC J01CR02), JemmaPass walks the ATC hierarchy up to J01C* (penicillins), matches it against Kurodo's allergy, and triggers the red alert. The patient summary itself is the safety check.
🐄
Haru
Haru
The Japanese grandmother · 春おばあちゃん
🇯🇵 Japanese national 👵 80 years old 🗣️ JA only 💊 Polypharmacy

Background. 80 years old, lives alone in rural Aomori. Mild congestive heart failure, seasonal allergies (treated with Allegra FX), occasional bronchitis (treated with Medicon Pro dextromethorphan). She speaks only Japanese, has no smartphone proficiency. Her grandson encoded her IPS profile and printed it as a Pocket Pass she keeps in her wallet.

Why she matters in the demo. Haru represents the "paper IPS" use case — for people who can't or won't use smartphones. Her printed badge is scanned by Kamekichi's app and by an iPhone using the native Camera app, proving JemmaPass interoperates with non-JemmaPass devices via its multilingual text channel.

📄 Haru's _j 1.2 profile

Raw: 1009 bytes Pruned: 777 bytes Wire: 500 bytes (~50%)
{
  "_j": "1.2",
  "p": {
    "gn": "Haru",
    "fn": "Tanaka",
    "gs": "F",
    "bd": "1946-02-08",
    "nat": "JP",
    "bt": "O+",
    "ct": ["+81-90-XXXX-XXXX"]
  },
  "al": [
    {
      "c": "419474003",
      "s": "L",
      "st": "A",
      "d": "Allergy to soy",
      "m": "Mild GI symptoms",
      "d_display": "大豆アレルギー · Allergie au soja"
    }
  ],
  "md": [
    {
      "c": "R06AX26",
      "t": "Allegra FX (fexofenadine 60mg)",
      "r": "O",
      "v": "1",
      "u": "tab",
      "rs": "BID",
      "rc": "J30",
      "d_display": "Antihistamine · 抗ヒスタミン薬"
    },
    {
      "c": "R05DA09",
      "t": "Medicon Pro (dextromethorphan)",
      "r": "O",
      "v": "1",
      "u": "tab",
      "rs": "PRN cough",
      "d_display": "Antitussive · 鎮咳薬"
    },
    {
      "c": "C03CA01",
      "t": "Furosemide 20mg",
      "r": "O",
      "v": "1",
      "u": "tab",
      "rs": "Daily",
      "rc": "I50",
      "d_display": "Diuretic for heart failure · 心不全用利尿薬"
    }
  ],
  "cn": [
    {
      "c": "I50.9",
      "d": "Congestive heart failure",
      "st": "A",
      "d_display": "うっ血性心不全 · Insuffisance cardiaque"
    }
  ]
}
Why this matters: Haru's full IPS — patient demographics, allergies, three medications, one condition — compresses to 500 bytes. That's small enough to fit in a single low-error-correction QR code, scannable by ANY phone camera with the multilingual text fallback channel. Even an iPhone with no JemmaPass app installed can read her warnings.
🐢
Kamekichi
Kamekichi
The mascot · The rescuer · 亀吉
🐢 JemmaPass mascot ⛑️ Disaster volunteer 🗣️ JA · multilingual UI 📱 Rescuer Mode

Background. Kamekichi (亀吉, "lucky turtle") is the brand mascot AND the rescuer protagonist of the demo scenario. A disaster relief volunteer in Shikoku, he's not a doctor — just a regular citizen with the JemmaPass app installed. The turtle motif comes from the Silicon Shell metaphor: the back shell holds the AI brain (Gemma 4), the front plastron displays the patient's QR.

Why he matters in the demo. Kamekichi demonstrates the Rescuer Mode — the most critical use case. He scans Haru's printed badge, receives Kurodo's SOS over Nearby (no internet), and uses the @Tool drug-allergy cross-check to detect the Augmentin × penicillin contraindication. He proves JemmaPass democratizes clinical safety for non-medical first responders.

⛑️ Kamekichi's rescuer capability snapshot

Mode: Rescuer Tools: 49 @Tool Network: Airplane mode
📡
Receives SOS broadcasts
JemmaNearbySosService.kt · Strategy.P2P_CLUSTER · 6 chunk types (V/VR/S/SR/E/F)
📷
Scans patient QR badges
ML Kit Barcode Scanner · supports _j2: direct payloads (no manifest fetch)
🌐
Auto-translates patient profile
Deterministic lookup via ips_valuesets_translations.lang · sub-millisecond
💊
Cross-checks medications
checkOneDrugAgainstFocusProfile @Tool · 302,516 DDI records · < 50 ms
📻
Records voice notes
Gemma 4 multimodal audio input · 16 kHz PCM + 44-byte WAV header
🚨
Triggers red alert UI
Full-screen TTS in patient's language · haptic feedback · no cloud roundtrip
✦ The Intersection

When their paths cross

On the day of the M 7.6 Shikoku earthquake, all three personas connect within 20 minutes — without any internet, cell tower, or cloud server.

1
00:00 · Quake hits

Cell towers collapse. Kurodo is injured. He activates SOS Mode on his Pixel 9 (a double-tap on the lock-screen widget — no unlock, no app launch). His JemmaPass begins broadcasting his pruned IPS via Nearby Connections.

2
+04:00 · Discovery

Kamekichi, 800 m away, sees Kurodo's SOS appear on his radar UI. The full IPS profile arrives via 5 BLE chunk reassemblies. Profile auto-renders in Japanese.

3
+08:00 · Detour

On the way, Kamekichi encounters Haru. He scans her printed Pocket Pass with ML Kit. Her IPS appears instantly. He scans her medication box → green check.

4
+15:00 · The killer moment

Kamekichi reaches Kurodo. He's about to give him Augmentin from his vest. He scans it first → red alert: beta-lactam cross-reactivity with penicillin allergy. Anaphylaxis averted.

Zero internet packets exchanged. Zero cloud server contacted. Zero medical training required of Kamekichi. One life saved — by 200 ms of on-device clinical reasoning.

Want to see this in action?

The personas you just inspected appear verbatim in the 3-minute demo video. Watch them come alive.

🎬 Watch the demo 📱 Try it yourself ← Hub